A New Purged UV Spectroscopic Ellipsometer to characterize 157nm nanolithographic materials

  • PDF / 1,652,561 Bytes
  • 6 Pages / 417.6 x 639 pts Page_size
  • 71 Downloads / 151 Views

DOWNLOAD

REPORT


177

Mat. Res. Soc. Symp. Proc. Vol. 584 ©2000 Materials Research Society

DESCRIPTION OF THE PURGED UV ELLIPSOMETER Two main differences appear when we want to work at 157nm. First, because of the strong molecular absorption bands in 02 and H 20 at 157nm, the entire beam path must be free of this kind of contamination. Second, the standard optical path used at SOPRA with a monochromator just before the detector and connected to the analyser arm of the ellipsometer by an optical fiber cannot be used. These two problems have been solved using the following solutions:

Figure 1: General view of the SOPRA PurgedUV SpectroscopicEllipsometer. SnectroscoDic ellimsometer setun The double monochromator is included in the polariser arm just after the deuterium lamp. This mounting ensures an optimized straight light rejection with a minimized beam path. The light beam goes through a MgF 2 Rochon polarizer mounted on a stepper motor. The reflected beam passes through another Rochon analyser and is detected by a photomultiplier in photon counting mode. The two arms are mounted on a high precision goniometer. The angle of incidence can be changed automatically in the range 7-90°. The system works in rotating analyser mode to avoid parasitic polarisation due to the monochromator. Polarisation sensitivity of the detector is calibrated in straight line. The spectral range is 145-300nm but it can be extended in the visible range. In addition to ellipsometry, the system can make photometric measurements at fixed polarisation state (reflectance and transmittance). Scatterometric measurements are also possible.

178

Purged glove box The entire system is installed inside a glove box with continuous H2 0 and 02 purification. Dry nitrogen is injected continuously in the box with automatic adjustment of the surpressure. The filters can be regenerated automatically every 3-months. One working face with three gloves allows adjustment of the different parts of the system, to mount the sample on the sample holder and to replace the deuterium lamp when needed. Samples up to 200mm diameter are introduced using a load lock. Residual H20 and 02 are measured continuously. They are in the part per million ranges during normal working conditions. 0.5 0.45

-o-157nm - --193nm

0.4 0.35

227nm --- 262nm -•-298nm

0.3

.i

0.25

I-

0.2 0.15 0.1 0.05 40

45

50

55 60 Incidence angle (deg)

65

70

65

70

I

0.8 -'157nm 0.6 0.4 0.2 6

0 C.

0

-a-193nm 227nm -,-262nm --- 298nm

-0.2 -GA

-0.6 -0.8 -1 40

45

50

55

60

Incidence angle (deg)

Figure 2: Experimental variable angle SE measurement made on a puropsil substrate.

179

EXPERIMENTAL RESULTS Only variable angle spectroscopic ellipsometry measurements on substrates and thin film on

substrate samples are presented here after. We have selected CaF2 and puropsil (natural silica) substrates and more unusual LaF 3 layers on CaF2 substrates. Puropsil substrate Variable angle SE measurements have been made at five different wavelengths (157.6, 193,

227, 262 and 298nm). The measurement is made f